NASA Johnson Space Center
Oral History Project
Commercial Crew & Cargo Program Office
Edited Oral History Transcript
Kurt D.
Eberly & Michael R. Pinkston
Interviewed by Rebecca Wright
Dulles, Virginia – 3 June 2013
Wright: Today is June 3rd, 2013. This oral history interview is being
conducted with Mike Pinkston and Kurt Eberly at the Headquarters of
the Orbital Sciences Corporation in Dulles, Virginia, for the Commercial
Crew & Cargo Program Office history project. Interviewer is Rebecca
Wright, with Rebecca Hackler.
Thank you very much for both of you being here. We know that you’re
key members of the Antares team, and we would like for you to help
us understand more about the work that you’ve done. Kurt, if
you would start, just give us a brief background on your time with
Orbital and your involvement with the program.
Eberly:
I joined Orbital on September 16th, 1991. The day will live in history.
Pinkston:
It will now.
Eberly:
It was straight out of college. Antonio [L.] Elias, one of our Vice
Presidents, was recruiting at MIT [Massachusetts Institute of Technology,
Cambridge]. That’s where I graduated from. I thought it sounded
like a neat place to work, since they just showed a video of the Pegasus
rocket having launched to orbit on its first try. I think that was
the first commercially-developed rocket at the time.
I thought Orbital Sciences must be a pretty neat place to work, so
I came on board in September of 1991 and worked Pegasus for probably
eight or nine years, and ended up moving to Vandenberg Air Force Base
[California] where we do a lot of our integration and launches for
Pegasus. That’s where we build up the rocket, and we actually
take it around the world and launch it from different regions. Then
I worked Minotaur, which is another one of our rockets, and then I
got involved in the missile defense project, Ground-Based Midcourse
Defense [GMD]. Actually, Mike and I worked on that for a number of
years in the early 2000s.
Then I had a cushy job as a Systems Engineering Director, hiring and
training and so on, supplying systems engineers to the various programs
at Orbital. In 2007 I was tapped to help lead what was called the
Taurus II rocket at the time. That was an idea that Orbital had, and
we can talk more about that later certainly. April 2007 is when we
started what was called the Taurus II project that became the Antares
project, and we just had our first launch in April of 2013.
Wright:
That was a good day, wasn’t it?
Eberly:
Sure was.
Pinkston:
That was a great day.
Wright:
Mike, could you share with us your background?
Pinkston:
Sure. I graduated from Oklahoma State [University, Stillwater] in
the summer of 1987, and I started with a company called Space Data
Corporation in Tempe, Arizona. That was a small company. I was attracted
to Arizona to begin with, so that was prime reason number one, I wanted
a job in Arizona. Then I toured on my interview and they were working
on a couple of what I thought were cool Space Shuttle-type payloads
so I thought, “I’ve got to jump into that.”
I did some small pointing systems for the Space Shuttle, a lot of
small rockets, and then around the [1988] timeframe Space Data was
acquired by Orbital. So I consider myself an Orbital employee for
my entire career, although I did start at a small predecessor company.
I’ve been essentially in the same place my whole career, since
1987. The cool Shuttle payload that I saw when I did my interview,
I actually ended up working on that one. Then another one after that,
and then a couple of small rocket programs.
I ended up also for about a nine-month stint at Vandenberg Air Force
Base working on Pegasus. That’s where I first met Kurt, back
in the early-to-mid-’90s. I worked Pegasus for a couple of years,
then as chief engineer on the Minotaur I program, through the development
of that product line. Then I got vectored over in 2000 to work the
proposal effort for what became our GMD interceptor program, which
again, I worked with Kurt for a long time on that. I worked the GMD
program for a decade, actually more than a decade if you count the
proposal. In 2004 I became the program manager, and then ran that
program through the end of 2010.
I spent a couple of years working a targets program for Missile Defense
Agency [in the Department of Defense], and then was tapped on the
shoulder in the summer of last year and asked if I would be at all
interested in taking over for the Antares Program Manager Brent [R.]
Collins, who was retiring. In fact, they had coerced him into staying
longer than he wanted to stay, and eventually Brent passed the baton
in October of last year. I’ve been here ever since. That’s
it in a nutshell.
Wright:
That’s a good nutshell. Let’s go back to when you first
learned about the COTS [Commercial Orbital Transportation Systems]
development, and the partnership between Orbital and NASA for this
venture. Can you tell us your thoughts, when you heard about how this
was going to be possibly different from what you’d worked on
before?
Eberly:
Yes. A little bit of background—Orbital had found somewhat of
a niche market in developing national security spacecraft, because
we build spacecraft as well as rockets. We had started launching several
developing satellites quickly and launching them fairly cheaply on
a Delta II rocket, which is a medium-class liquid rocket that’s
still in service, although it’s being phased out. As it was
being phased out, the writing was on the wall. The Air Force ended
its support of Delta II because its mainstay was launching GPS [global
positioning system] satellites, and they decided they were going to
launch GPS satellites on EELVs [evolved expendable launch vehicles],
larger rockets. It was pretty clear that Delta II was not going to
stay in production.
The Orbital management was excited about this recent business, which
was developing these low-cost national security spacecraft, and then
launching them on this medium-class rocket. If that medium-class rocket
went away, that market would no longer be there. The first idea for
Taurus II was as a replacement for Delta II, so that we could continue
this market. So it actually predated the COTS effort. We had done
some studies internally within Orbital for years, around the 2005
timeframe.
Then, in January 2006, the COTS RFP [Request for Proposal] was issued.
We were excited. This was a corollary market that would also help
us to bring the resources to bear to develop this new launch vehicle.
We submitted a whole proposal, and the spacecraft we were bidding
at the time I believe used Russian proximity operations avionics.
We thought that was a low-risk approach because it was up and running
and the Russians already had automated rendezvous in place. And we
were not selected. SpaceX [Space Exploration Technologies Corp.] and
[Rocketplane] Kistler were selected by NASA later in 2006, so we just
kept the Taurus II in study phase at that point. I think our CEO [chief
executive officer, David W. Thompson] and our Board of Directors needed
more assurance that there was going to be a strong market.
Then Kistler was terminated in October 2007, a year and a half later,
and the funds that were remaining on the COTS program from Kistler
were re-competed. It was a new RFP that was issued, so we were very
excited to have an opportunity to have a second go at it. We think
that part of the reason we weren’t selected was we had the Russian
docking hardware, although that was never really explicitly stated
in the RFP documentation.
So we were careful this time to have a different approach for our
spacecraft, and we elected to go with the Japanese proximity operations
avionics which were built by Mitsubishi Electric Corporation, the
same rendezvous approach system that the [Japan Aerospace Exploration
Agency] HTV [H-II Transfer vehicle] uses right now. We tried to present
a very reasonable approach to building the rocket and building the
spacecraft, drawing on heritage from our other launch vehicles.
Also, because it’s a much bigger rocket than we have typically
flown on Pegasus or Minotaur or Taurus, we decided to go with a liquid
first stage, which is typical for rockets of this class. We decided
to hire a Ukrainian company, Yuzhnoye [Design Bureau], and their sister
company, Yuzhmash [A.M. Makarov Yuzhny Machine-Building Plant], to
build the first stage core, the tanks, and the pressurization system.
We think that that brought a lot of heritage from the Zenit rocket
program.
The upper stack—we decided to go with a solid rocket motor,
which is very similar to how we launched a lot of our other rockets.
We buy solid rocket motors from ATK [Alliant Techsystems, Inc.] and
we put Orbital avionics on them. We analyze it, we do all the software,
we do all the structures, and we do all the integration work. From
the first stage up it was very similar to past rockets that we had
done. That’s the rocket.
On the spacecraft we planned to use a bipropellant propulsion system
that was very similar to a lot of the large geosynchronous spacecraft
that we build and market commercially. We hired Thales Alenia Space
in Italy, who are the makers of the MPLM [Multi-Purpose Logistics
Module] which used to fly inside the Shuttle as a logistics carrier.
We hired them to do that same basic job for Cygnus. We think that
pulling on that heritage from these suppliers that had done this kind
of work before presented an attractive proposal, in terms of low risk
to NASA, and that we’d be able to execute the development.
We were selected in February of 2008, so that really gave our CEO
and the Board the impetus to move ahead with this project. We [Taurus
II group] were funded starting in April of 2007, so roughly a year
prior we had started the first phase of development under internal
research and development funds. That’s Orbital funds. The COTS
award, which at the time was just one single launch of the rocket
and the spacecraft that were going to demonstrate that capability
to go to the [International] Space Station [ISS]—winning that
assistance in the development and that purchase of a single demonstration
mission gave the assurance that we needed to our CEO and our Board
to free up additional funds to fund through to the first launch. That
was a really big step.
I think that’s the crux of the COTS and the CRS [Commercial
Resupply Services] development paradigm. The government says, “We’re
going to promise that if you develop it, here’s a little bit
of money to help you,” but certainly it’s not going to
cover everything or even half. “If you then put in the rest
of the money and get all the way through development, we will buy
X number of launches from you.” In the case of the first COTS
award, it was just one single launch of an Antares rocket and a Cygnus
spacecraft.
The government caps their risk by saying, “We’ll pay you
by milestones. When you get to PDR [Preliminary Design Review], we
might pay you some money. If you do a certain development test that
retires risk, you test-fire the first stage, we’ll pay you a
little bit more money,” and so on. And the same with the spacecraft.
They limit their risk and their exposure to the commercial development,
because what if we ran into problems or we had cost overruns? That’s
all on the contractor to cover those kinds of cost overruns, and the
government is able to limit their risk, their exposure, for the development.
That’s really what’s new about the COTS Space Act Agreement
method of developing new hardware. I think given the amount of money
that’s been invested in COTS—I think it’s about
$800 million between SpaceX and Orbital, and the amount that was wasted
on Kistler—for $800 million you get two new launch vehicles
available to NASA, two new maneuvering spacecraft capable of rendezvous
with the Space Station, and two new launch pads that are available
for launches to the Space Station, on two different launch ranges.
I think that that’s an amazing success story for the government,
in terms of developing two new systems that are pretty complicated
and have some stringent requirements.
Wright:
And different. In contrast to what SpaceX chose for their route, your
route was one that you thought was low-risk, wrapping in the international
collaboration with their proven heritage components.
Eberly:
Yes, that’s right. We definitely have a different approach.
We tend to subcontract out more of our systems, and I think SpaceX
tends to try to do everything in-house that they possibly can. There
are certain things we do in-house, avionics and harnessing and software
and guidance, but we typically will go out of house for a lot of our
big subsystems.
Wright:
NASA provided some of the funding. What other aspects do you feel
that NASA contributed to the development of the Antares?
Eberly:
They made available expertise in certain technological areas that
are very hard to get. We had a problem with the AJ-26 engines that
we selected for the first stage. They’re the original Soviet
moon rocket engines that were built for the Soviet moon rocket, the
N-1, and they were placed in storage when that program was cancelled.
They were built in the ’70s, so they’ve been in storage
for a long time and they have materials on them that are susceptible
to stress corrosion cracking. It was not known to us that this was
the case, nor to the vendor, Aerojet.
Aerojet purchased these from a Russian supplier in the ’90s,
and they had about 40 of them in Sacramento [California] that they
sold to us as the AJ-26 configuration for Antares. This fundamental
flaw in the materials—they were never designed to be in storage
that long—was found in an early development test where we fired
the engine at [NASA] Stennis [Space Center, Mississippi].
The expertise that NASA brought to bear on those engines, they really
helped us through the process to where we can now identify the defects
in this material. We can screen it out, we can do weld repair, and
then have a flightworthy engine after that process. NASA brought their
Marshall Space Flight Center [Huntsville, Alabama] materials expertise
and lab to bear. I think it was essential to bring all that history
that they have working with cracked metals on other vehicles. They
certainly have a lot of experience and a lot of lab equipment that’s
very costly and expensive. They’re able to really step in and
quickly analyze the materials and buttress the claims of Orbital and
Aerojet, and help us through the process of developing the weld repair
process and the inspection process that assures us that the engines
are going to be flightworthy when we fly them.
Wright:
Did you sense a partnership of working with NASA personnel?
Eberly:
Yes, absolutely. Our strategy was to be completely open with them
and treat them as investment partners, which was how they wanted to
be treated. The COTS office itself was very small, it was really just
two or three people. When they wanted to provide this help, they had
to reach out to the various Centers where the expertise resides. It’s
not like they had a whole program that was just managing us.
We kept them informed every step of the way, and we still do. Anything
they want to see, we show it to them. I think there’s only about
one meeting that they’ve ever requested to attend that we’ve
said that we’d rather have that internal. We’ve tried
to keep them fully informed of our progress, every step of the way.
In return, they’ve come forward and when they hear a problem
they say, “Hey, we know of this expertise, it resides here.
Can we apply it for you?” They will pay through some funds that
they have available for risk reduction, risk management, and they
funded the support to us in most cases where it’s development
work.
There’s also another example where we had an issue about contamination,
particles in the propellant storage tanks on the core, and whether
that presented a risk to the rocket engines in terms of an unintentional
ignition. We use liquid oxygen as the oxidizer, and so metal particles
can strike another metal surface in the engine and create a spark,
and cause an ignition. There’s an expert at [NASA] White Sands
[Test Facility, New Mexico], Joel [M.] Stoltzfus, who we asked to
help us. He came in and they funded him.
He brought some test data that really only he is going to have. They’ve
done these studies, firing metal particles at various angles against
a metal plate in a LOX [liquid oxygen]-rich environment and seeing
if an ignition occurs. That kind of work is painstaking, it’s
hazardous, and it’s difficult and expensive. He was able to
bring that experience which had been developed for many other programs,
including Shuttle, this database of knowledge. That was really helpful
to us to get past that issue and get through the first launch. That’s
a couple of examples of the things that they’ve helped us out
with.
In terms of Stennis, we needed a place to do the acceptance testing
of these engines and practice. The great thing about a liquid rocket
is you can fire them as an acceptance test before you launch your
rocket, and then turn them off and clean them up and fly them. You
have a very good assurance that they’re working properly, as
opposed to a solid rocket motor, where you just ignite them once and
there’s no way to shut them off. We were looking for a test
stand for our engine. So there we were a commercial user on a NASA
Center. If there was a commercial liquid rocket engine test stand
somewhere we could have gone there, but it turns out that these are
some of the specialized facilities that are very hard to find in the
commercial world. You can’t really go down to Joe’s Rocket
Test Stand Services and hire them.
That’s where we came in as a commercial user. We had a Space
Act Agreement directly with them, a separate reimbursable Space Act
Agreement, where we paid them. Instead of the COTS one where NASA
was paying us, this was a different one where we were paying Stennis,
and still do. We go down there with an engine and we pay their time
and materials. We’ve gotten a lot of support from Stennis. We
pay the direct costs that we incur, but they bring this history of
testing rocket engines and configuring the test stand, and getting
the propellants to flow correctly to the engine so you don’t
have any issues. They reconfigured the test stand from a horizontal
to a vertical configuration, and they put the instrumentation in them.
That’s the kind of stuff that we’re able to tap into those
decades of experience.
Pinkston:
Comparing the cost of duplicating that capability ourselves, it’s
just an enormous advantage.
Wright:
And the delay of finding someplace that’s readily available.
Eberly:
Right.
Pinkston:
The other area where NASA’s really helped out is in the development
of the fueling facility. That was another area where, as I was coming
onto the program last year, we ran into some significant challenges
with getting that facility operational. Both in terms of people, and
material that NASA had in inventory as residual that they were able
to provide to us. There were a number of examples we can provide where
people and hardware and expertise were all brought to the table and
helped us over that obviously critical hurdle in terms of getting
the facility up so that we could get into the integrated testing late
last year.
Wright:
And this was at MARS [Mid-Atlantic Regional Spaceport, Wallops, Virginia],
is that what you’re referring to?
Pinkston:
Yes, that’s exactly right.
Wright:
Would y’all talk about that part of the partnership? Because
it’s not just a partnership with NASA, but now you’ve
got another very vital part of Orbital’s future there.
Eberly:
Sure. MARS is the Mid-Atlantic Regional Spaceport. It’s primarily
[the state of] Virginia that’s the major funder, but they do
coordinate with Maryland. Wallops Flight Facility is part of [NASA]
Goddard Space Flight Center, which is in [Greenbelt] Maryland, and
the good Senator from Maryland, Barbara [A.] Mikulski, is a big supporter
of NASA. She’s on the appropriation committee that deals with
the NASA budget. Goddard is in her district, so she’s taken
a really keen interest in the project—COTS in general—and
she’s a big supporter of commercial space, in particular Wallops
and Goddard. She helped to make sure that the project was supported
properly politically, and that’s been a big benefit.
The spaceport has been supported by Virginia, and it had been a part
of Virginia since I think the mid-’90s. We’ve done some
commercial Pegasus launches out of there before. They provided the
launch pad for the Minotaur [rocket], and we launched four or five
Minotaurs prior to this year off of Pad 0-B. But it only had about
four or five employees, and they’d never done anything this
big. Really, this was a big step up for them to grow their capability
to have a launch pad. The idea was that they had bonding authority,
so they could issue bonds on the behalf of this project and raise
the funds to help offset the total cost, and we would contribute funds
as well.
The idea was to build something akin to an airport, where they would
build the airport and operate it, and then you’d have root users
like airlines come in and say, “I want to do a launch and I’ll
pay you a user fee, an occupancy fee.” That’s the idea,
and the benefit to Virginia is that economic benefit of bringing programs
and users to the Commonwealth of Virginia. The [Virginia] Economic
Development [Partnership], under the Governor’s control, also
helped to sway the decision to come to Wallops Island and to MARS.
We had a decision that we had set up between Cape Canaveral in Florida—Space
Florida was a state organization that was trying to entice us to come
there, and MARS was the equivalent for Virginia. Ultimately Virginia
won out for a number of reasons, including they just had a better
plan with the bonding authority to raise funds to get the project
done, and we’re located in Virginia and we thought the proximity
was a little better. Plus, Cape Canaveral, the Eastern Range tends
to get a little crowded sometimes with the EELV launches going out
of there. The thought was that we’d have a clearer launch schedule
going out of Wallops, and be able to launch when we really wanted
to and when NASA wanted us to. For those reasons we chose to go to
MARS.
Having said that, they razed the old Pad 0-A, where the first and
only Conestoga rocket had been launched. They took it down and then
we built it up, brand-new. The project has been quite extensive. Certainly
was a bigger and more complex and more difficult project than any
of us had entertained. There’s just a lot of complexity with
what’s called a liquid fuel farm, which is the storage tanks
full of equipment, full of liquid and gas, that you have to load into
the vehicle to get it to fly right. Getting that all supplied to the
vehicle at the right cleanliness levels and temperatures and pressures
and flow rates is very, very challenging. That’s something we
certainly underestimated as a company.
We’re through it now, with a lot of help from Virginia, from
NASA, from Wallops. [NASA] Kennedy [Space Flight Center, Florida]
sent folks up, NASA Stennis—all through JSC, helping us out
by giving us additional resources. They really came through for us.
We’re done, it works, it worked great for the last launch, and
we have a lot more knowledge now of what it takes to get one of these
medium-class liquid vehicles launched.
Wright:
Tell us about the magic day in April [first Antares test flight].
Eberly:
April 21st, 2013.
Pinkston:
The day that shall live in infamy.
Eberly:
It was our third attempt to launch the test flight. The first attempt,
we had gotten very close but we had an umbilical prematurely disconnect.
That shut us down for that day, but we solved that problem pretty
quickly. That was on a Wednesday. We went again on Saturday, but the
winds were very strong at altitude, and they were in a very unusual
direction. The way we had planned for the rocket to fly resulted in
the potential for debris in the event of a destruct action, which
you always have to account for. It could have ended up on Assateague
Island, which was populated with some people.
We had to call an abort for that Saturday, much to everyone’s
disappointment, although I was a little bit relieved that we weren’t
going to try to launch with such unusually strong winds. The first
launch of a new rocket, you’re relying so much on models of
how it’s going to behave once it lifts off. You think you know
how it’s going to behave, but you could be wrong about certain
aspects of the guidance constants or the autopilot parameters. It
might be a little bit less stable than you thought, and an extra wind
gust or two could make a difference in terms of success or failure.
First launches of new rockets have a fairly poor success rate if you
look historically across the fleet in the U.S., or even internationally.
So it was very tense. The next day, on Sunday, it was predicted to
be calm and it was. A really good wind situation, and we just ticked
right down through the countdown. Got down to inside of three minutes,
and that’s when we start our auto sequencer. Everything looked
great on the data. The pad had been supplying propellants very repeatably
in the last two attempts to the vehicle, so we were all in the control
room, had the screens up with the various shots.
I think we all had our favorite view that we wanted to see dialed
in on the TV camera, and it was just great to finally—we had
done a stage one development test prior, called the 7000 test, where
we practice fired the first stage on the launch pad. We had gotten
down to half a second before ignition and had an automated abort.
We fixed that timing issue, and when we got counting down I was thinking,
“Oh my gosh, I hope this thing doesn’t abort.” I
saw the counter coming down, and I just turned my attention to the
TV screen and I was so happy to see flames coming out of the flame
trench.
There really weren’t as many flames as I thought, and this is
the first time we really had everything up and running to lift off.
We had a lot of water—you may have seen the large water tank
that’s adjacent to the pad, and that was to supply cooling water
to the pad. It really mitigated a lot of the flames. As it lifted
off, we do a maneuver that Antonio [L.] Elias called the [Paul] Baumgartner
Maneuver [named for the Lead Guidance, Navigation, and Control Engineer
on Antares], where you’ve got this transporter erector launcher
that’s rocked back.
You don’t want the flames to toast that thing because you want
to save it for the next launch, so first you kick the attitude this
way [demonstrates], and then the rocket goes that way, and the flames
never really impact the transporter erector launcher. We thought,
“The magnitude’s like a degree or two, you’re never
going to see it on the video,” but I think everyone saw it and
it was very exciting to see that happen.
Pinkston:
It was not exciting, it was scary.
Eberly:
It was scary. The other thing I really noticed was that it was so
slow in lifting off. Compared to a solid rocket motor powered vehicle,
it’s so slow coming off the pad. It just seems to take forever.
Then I calmed down. As soon as it got away I thought, “This
thing is flying straight and looks great.” It was 235 seconds
of the first stage burn, and that’s a long, long time that these
rocket engines—which we’ve had problems with. We had the
cracks, we had the test failures. We’d done our homework, but
you can never be sure that everything’s going to work properly
together. All the dynamics involved when you lift off, the vibrations
and the aerodynamic model of flying through the air and so on.
But it was just perfect. Got through the first stage burn, burned
out, separated the stages. By then I think we thought we were kind
of home free. People were clapping in the control room when the first
stage burned out and then MECO [main engine cut-off] was announced,
and then that separation—everyone clapped spontaneously. You’re
supposed to hold your fire until the payload is separated successfully
in orbit, but I think everyone’s so relieved that that part
of the mission was done that they couldn’t resist. It was a
happy day.
And all the data review after the flight has shown that it was really
just flawless in terms of the execution. The models that we were so
worried about—the slosh model on the first stage, the thrust
frame bending model, the autopilot design—everything was so
spot on. It really shows that we have taken the A-Team from within
Orbital. We really had our best people up and down the line and in
every position. This is a very important project for Orbital Sciences,
so we were able to get those people applied to this project, and show
that we really nailed the data that we’ve reviewed. We just
had our final data review, two days of data review, last week to our
customer.
Wright:
Yes, I had read that once you reviewed the data it really was as good
as you thought it was.
Pinkston:
Yes. I’ll just tack on a couple of things to what Kurt said.
He mentioned the modeling aspect, and that’s always a huge source
of uncertainty in a first flight. I think a large reason why the first
flight’s success rate, if you look across the history of launch
vehicles—depends on the bar you set based on what you expect
in a recurring launch program—it’s not very good. You
multiply that for this one for us, it being our first large liquid
rocket, and a real eye opener to me is just how much more complex
and particular the guidance and control of the first stage is in a
large, lumbering liquid rocket.
We were very conservative in the wind constraints we set, both to
avoid banging the rocket into the launcher, as well as some of the
upper level winds, both of which caused us to abort on Saturday. It
also caused us to wave off even an attempt on the Friday before that.
To go through that first flight and to go through the data and see
everything really lined up in spades with what the models were predicting—it
was really satisfying I think for the whole team.
A lot of folks worked a long time on this program, and I think frankly
that’s another area where having some independent expertise
on the NASA side—I heard nothing but rave reviews of the independent
verification that was done on our guidance and control system that
NASA had provided. It really bolstered our confidence going in, and
I think the results bore out just how critical that help was to us,
in terms of getting it right the first time.
Wright:
Can you give us some examples about the things that you learned through
the testing? You mentioned the firing test and the cold test, the
different phases that you went through that you were able to learn
what you needed to know.
Pinkston:
It started all the way back last spring, early summer timeframe, when
we had just started doing some of the initial testing of the fueling
facility itself. We learned a number of things in that process, a
lot of lessons learned. There were a lot of headaches that had to
be worked off. We actually found one of the key subsystems, the chilled
helium subsystem, was not coming close to meeting our requirements.
That is a very notable area where NASA was able to come through. They
found a residual heat exchanger that they had laying in that bone
yard somewhere that we were able to bring out between that 7000 test
and the first test flight. We cut that in and fixed that problem for
good. You learn how to get the facility to flow commodities.
When we first brought the test article up, which is essentially a
full-up first stage with some elements of the flight avionics on it
to make it do what it needs to do to load propellants and fire the
engines, some of the initial tests there were disappointing in the
results. When you start seeing the interaction of pressurization valves
on the core and how they affect the fueling facility, we had a number
of problems there that we had to work our way through.
We got to the point where we were confident that we could operate
the system without blowing relief valves on the facility side, and
then we got into what was called the 5000 series testing, which was
a set of incremental tests where we started with loading liquid oxygen
on. Then we advanced to a full load of liquid oxygen and some chilled
helium, and the third test was when they’ll actually put kerosene
and the full load of chilled helium on it as well. You kind of work
your way into it.
You’ve got three specific tests before the hot fire, and we
had to run each of those tests twice. Each time we learned something.
I don’t think I remember a case—after the initial bringing
out of the system, I don’t think we ever had a problem with
actually getting commodities on or off the vehicle. But we found all
kinds of other things—timing interactions between the loading
system and the flight system. They’re both software-driven systems,
and you find little timing idiosyncrasies that abort you out at the
wrong time.
Each of those tests we learned something. We learned something important
and we folded in a change, folded in the change and then got to the
hot fire event. First attempt at that, as Kurt said, we aborted out
at somewhere close to half a second before the engines fired. It’s
just devastating. You’ve been working to get that close, but
we figured out what went wrong and we fixed it and tried again, and
that test was great. We got a lot of really good data off of that
hot fire test.
Another pretty amazing thing, to me anyway, is that we had almost
to the day two months between the 7K [7000] hot fire test and the
actual test flight launch. We did the hot fire on February 22nd, and
we did the launch on April 21st. In that two months we completely
refurbished the pad after the hot fire test. In the hot fire test
itself we measured lots of environment instrumentation on the first
stage and the first stage components, and we found some things exceeded
our expectations.
We had to go back and re-qualify some components, got the test article
off the pad, and another little quirk is that we’re actually
going to reuse all that hardware. There’s some on-pad cleaning
operations and servicing that we had to do that was significantly
hampered by weather. We were able to get all that done, test article
off, get the pad refurbished, re-qualify components that need to be
re-qualified, reviewed all the data, went through all the requisite
reviews in a two month time period. It’s really quite amazing
to me that we were able to sweep through it.
Wright:
Busy.
Pinkston:
It was extremely busy. I think the short answer to all of that is
we just learned a ton. I mean, from the time a year ago when we first
started testing the facility, through the 5K testing and the 7K testing,
to the test flight—it was just a constant acquisition of knowledge
that allowed us to really trim up the operation. Again, I think the
results of the test flight showed how well that worked.
Wright:
Take us now from April 21st to the next upcoming demonstration flight.
Pinkston:
We’ve got a lot of hardware, so that’s not a huge problem
for us. I’ll say that over the course of the first third of
the year we were pretty squarely focused on 7K and the test mission.
So we did get ourselves a little bit behind on processing the rocket
for the COTS demonstration mission. We were trying to hit a June date,
which would have been another close to two month turnaround. One of
the things everybody was quite concerned with was how does the pad
come through a launch operation? If you look at the condition of the
launch pad after the test mission flight, it was really in outstanding
condition compared to what it could have been, so the refurbishment
of the pad is going to be fairly straightforward.
There are some changes we need to make, there’s always a few
lessons learned. Pieces that fly off of things that shouldn’t,
that need to be secured a little better. We burned up some cables
that we wish we hadn’t, so we’re off repairing those and
replacing a few. I think all in all we’re going to have a pad
that’s ready by the mid-to-late-July timeframe. That’s
good. The rocket, we made a tough decision to actually replace an
engine.
As Kurt mentioned previously, these are aged engines and the one that
we had out there for the next flight, there’s a minor defect.
We worked with Aerojet and the original Russian manufacturer, and
they’ve all given us a clean bill of health on it. We had some
residual questions that just hadn’t quite been answered, so
we made the decision to go ahead and push that down the line a little
bit and give ourselves more time, just to make sure we’re absolutely
certain we understood the conditions and not carry any extra risk
that we weren’t ready to carry.
I think very soon we’ll get past that. We made the decision
almost a month ago to go ahead and swap that engine out. That delayed
us into probably a mid-to-late August readiness for the COTS demo,
and there’s a conflict at this point with the Space Station.
I think there’s an HTV mission that’s currently slated
in early August. That’s still not completely shaken out, but
that probably pushes us into a launch that’s late August, early
September by the time everything shakes out.
But I think we’re making good progress. Short form, I think
the pad came through in really good shape. We’ll get it turned
around and it’ll be ready late July. We made the decision to
swap the engine, and that put probably another four weeks of risk
into our schedule. That pushed us out into late August, mid-to-late
August. We were probably looking at a mid-to-late July readiness before
that, and with the HTV mission right there it was likely we weren’t
going to have the window that early anyway.
Wright:
Who would have thought, got to take a ticket. Rebecca, do want to
ask some questions?
Hackler:
Sure. You mentioned that last week you had the two-day review. Can
you talk a little bit more about not only that milestone, but how
you worked with NASA in general to verify that you had met all the
requirements for milestones. Did those require any negotiation to
ensure those were completed and you could get your payment?
Eberly:
I would say, as we’ve talked about, we’ve been very open
with our counterparts—Bruce [A.] Manners, Kevin [M.] Meehan,
and Alan [J.] Lindenmoyer—at the COTS office, and they’ve
brought in additional expertise as needed. The Space Act Agreement
was written in 2008, and here we are launching in 2013. Back in 2008,
before the rocket or the spacecraft were even developed you had to
write down, “Here’s the milestones we’re going to
achieve and here’s the payments that are going to go with it.”
The criteria are listed in the Space Act Agreement, which is akin
to a contract. We purposefully, on both sides, kept things a little
vague in terms of what were going to be the criteria. As we got closer
to each milestone we would sit down with Bruce and his team, and usually
they would take the lead and say, “Based on all we know about
your rocket, here’s what we expect.” A good example was
stage one assembly completion.
It sounds pretty easy, the first stage should be done, but there are
certain parts that don’t get put on the first stage until you’re
ready to launch. Ordinances that are hooked up, and pivots are not
removed, so there’s some things that seem pretty obvious, but
nonetheless you have to work your way through and get everyone to
understand, the first stage is semi-complete. We’ve got the
rocket engines, the thrust ring, the stage one core. We’ve put
them together and we’ve tested them. These other parts that
don’t go on until later, we agree they’re part of the
first stage, and they aren’t normally installed at the time,
but we’ll make sure that they’re accounted for and that
we have them in our possession.
We worked out a process where they came to Wallops and in essence
did a physical configuration audit where they walked down to the vehicle,
took pictures, and then we showed them all the build documentation.
We have a pretty complete set of build paperwork, just as part of
our normal process for each part. It’s tested properly, it’s
got the right pedigree, it’s the right part number and so on.
Then we show all that and deliver that to NASA, give them a copy.
It’s not a contract deliverable per se, but it’s the supporting
documentation that helps them know that we’ve met the milestone.
I think they had some oversight from the GAO [Government Accountability
Office] and from some of the members of Congress that were feeling
like, “Oh, these Space Act Agreements are a little loosey goosey
and they allow the contractors to get paid when the government can
go at risk.” I think they were feeling a little heat from their
management, from [NASA] HQ [Headquarters, Washington, D.C.] and from
GAO.
They had been audited, which I think is fine and we supported their
audits. As we came to understand it, it was good for us to supply
them with this kind of documentary evidence of completion of the milestones
because it helped them have an audit trail. It helped the program
stay supported, and to answer the critics that were trying to say
that this was a boondoggle of some kind for the commercial space industry.
I think they were reasonable. Once we explained to them, “This
is what we can show you for this particular milestone,” we may
have had a little back-and-forth, but we came to an agreement pretty
quickly. There’s never really been a dispute about whether we
met a milestone or not. We never go to them and say we’re done
when we’re not, and they know that because we keep them informed.
A lot of the milestones are pretty objective, and so it’s pretty
clear when they’re complete.
Speaking of milestones, we did add in the augmentation funds. When
we started the project it was really just the one launch of the rocket,
and as I said earlier, the first launches of a new rocket don’t
have a great track record. The thought was, “We don’t
have money for a dedicated test flight, so we’re just going
to have to put the first operational Cygnus on top of the first operational
Antares and launch it, and hope that we succeed and get all the demonstration
milestones completed that would allow us to proceed on with the cargo
delivery flights next.”
I think SpaceX started out with roughly $100 million more of the COTS
money to begin with, and then there was some decrement from what Kistler
spent. We started a little bit in the hole, so we couldn’t fit
in the multiple test flights that they had proposed. Everyone agreed,
both on the NASA side and the Orbital side, that having a dedicated
test flight of just the rocket before you put this high-value, first
of its kind Cygnus spacecraft on board would make a lot of sense and
retire a lot of risk from the program.
When funds were made available [in fiscal year 2011], I think we were
in complete agreement between us and NASA. The first priority would
be a test flight of the Antares, and then the construction and launch
of a dummy Cygnus. It basically had the same mass properties and shape,
but was heavily instrumented so they could measure the environments
that the real Cygnus would see, very precisely, on the ascent, because
the rocket launch is the most astringent environment that a spacecraft
will see in its lifetime.
That was all very successful. We collected a lot of data from the
Cygnus mass simulator that was flown on the [Antares] test flight,
and I think we’re very happy that we’ve retired a lot
of risk from this dedicated test flight. I think that was a really
key move for the program, to be able to fit that in.
Hackler:
As you progress through your development, because you are doing rocket
science, which is called rocket science for a reason, there were some
milestone slips and development fell behind the original projected
schedules. How did you address those schedule delays?
Eberly:
We had a weekly telecon [telephone conference], we gave scheduled
updates, we had a monthly scheduled working group at the JSC, and
we presented our best estimate of the schedule going forward. When
we added in the test flight we agreed to delay the first launch by
some amount, I think from December [2012] to March of the following
year. We were waiting for the pad to be complete, and we also had
the trouble with the rocket engines that we had to resolve, so that
really set us back.
We just tried to keep NASA informed every step of the way, and with
our best estimate of where we thought the schedule was going to end
up. We were wrong many times, in terms of how long we thought it was
going to take for both of those things to be resolved. To their credit,
they were patient. I think the additional Shuttle flight that they
got approved [STS-135] that carried up the last MPLM that was chock-full
of supplies—and I think they left it attached to the Space Station—that
bought them more time, in terms of needing the logistics train to
start coming down the pike. That was lucky for us, that they were
in a position where they actually didn’t need the cargo as early
as had been projected back in 2008.
Hackler:
You also worked a lot with the ISS Program Office, coordinating their
safety concerns and scheduling. Can you talk a little bit about your
relationship with that group of folks?
Eberly:
Have you talked to the Cygnus folks already?
Hackler:
No, we haven’t.
Eberly:
I think that’s who you should talk to because the IRD [Interface
Requirements Document] is really between Cygnus and the ISS. Cygnus
takes those safety requirements and flows them down to Antares. We
don’t really have to interface with the ISS at all. We drop
off Cygnus well below the Station, so there’s no collision hazard,
there’s no risk of any kind of problem. Then Cygnus, once we
separate, they maneuver up and do the rendezvous. So it’s really
on them to do all their really intricate fault isolation and safety
requirements that you have to meet in order to be able to approach
the ISS safely. That’s really not in the rocket’s job
jar.
In terms of the scheduling, for the test flight we weren’t going
to the ISS so we didn’t have to deal with that. As Mike mentioned,
we have a slot in September and we have a slot in December, and so
we are trying very hard to make sure we meet those. We understand
how crowded the schedule is with all the visiting vehicles, having
the right crews there with the right training.
Pinkston:
It’s a whole lot of moving parts.
Eberly:
Yes. We’re starting to really appreciate the difficulty of the
scheduling process for all the ISS partners and users.
Hackler:
Do you work with the FAA [Federal Aviation Administration] Office
of Commercial Space Transportation for the launch licensing?
Eberly:
That’s right, exactly. That was in the Space Act, that we would
be commercially licensed to launch, and the CRS contract says the
same thing. We’re responsible for all that regulatory work,
so we have a licensing lead. We’ve done a number of commercial
launches within our company, mainly on Pegasus, where we get the commercial
license. We’ve actually done a commercially-licensed launch
from Spain, we took the Pegasus over there. So we’re familiar
with how to do all that and stay on the right side of the ITAR [International
Traffic in Arms Regulations] rules.
That’s worked pretty well; the FAA has been pretty easy to work
with. The only thing I’ll say is there’s sometimes a disconnect
between the requirements that the FAA has and the requirements of
the launch range, and it’s just disappointing that they’re
not in sync because they really need to be. My understanding is that
the ranges can change their regulations more quickly than the FAA
can change their laws, and that’s why there’s that disconnect
at times. But I think they’ve gone pretty smoothly. They mostly
defer to the range safety folks for all the flight safety and ground
safety stuff, and the FAA folks are really in monitor mode.
Hackler:
I was thinking as you were talking about your backgrounds and your
years of experience on other projects at Orbital, that you had a [knowledge]
base you were ready to apply to this new project. Are there any specific
areas you can think of where that helped?
Eberly:
That’s right, yes. I think within Orbital Sciences and the launch
systems group that we’ve worked in for our careers, we have
become adept at flying new configurations of rockets because we try
to use all the same hard stuff. That is the navigator that senses
your position and your attitude, the flight computer that interfaces
with it and runs the autopilot and keeps the rocket flying in the
right direction, and the guidance that then sequences all the separation
events and the ordinance events. That set of avionics, and the software
that links them all together is done already because we fly it on
all of our other rockets. We really, really try to make a common implementation
of the avionics and the software and the guidance, which can really
take forever. If you didn’t already have that and have the analytical
tools to go and test it and verify that it’s working right,
that can take a long time.
So we were able to adapt that over from other vehicles that we built.
A lot of the targets that we’ve talked about for missile defense,
a lot of them are one-off configurations. They’re trying to
do a test of a radar, but they want a rocket that flies like this
and it has this aspect angle and this speed and this velocity and
this radar cross section. You end up taking strategic motors that
they’ve got in storage somewhere from a discontinued ICBM [Intercontinental
Ballistic Missile] program, and put together a new vehicle that has
three stages.
We constantly have to analyze the environments that are going to be
induced by a new configuration. What are the loads, what’s the
aerodynamics, and where does the autopilot need to be? I think that
has served us very well. “This rocket flies slow, it’s
got a very regular aerodynamic shape, it’s only got two stages”—that
kind of expertise and experience, and the analytical tools we’re
able to bring along really paid off for us I think.
Pinkston:
Just an example from my history—I talked you through what I’ve
done from the Pegasus program, the Minotaur I program, the GMD program,
and the IRBM [Intermediate-Range Ballistic Missile] target program
that I’ve worked have all used the same basic set of hardware.
The same rocket motors, with little tweaks and twists here and there,
same avionics. They evolve over time, with obsolescence and added
capability, but basically the same set of avionics—the same
navigator, the same cold gas attitude control system components. You
build an experience base with that set of hardware that’s just
invaluable. I come over to this program and I look at it, and, “Hey,
gee!”
It’s a hell of a lot bigger around, and it’s taller, so
our structures are bigger, but all the stuff inside is the same. On
that upper stack is a new rocket motor, it’s a bigger structure,
but everything else in there is stuff that we’ve done before.
I don’t think there’s any substitute for building that
kind of historical, heritage-based understanding of what you’re
applying. As Kurt said, all you’ve got to do at that point is
figure out what environment it’s going to see and make sure
it doesn’t exceed any of our qualification levels.
The real new thing on this was the first stage. I think our history
has always been finding the best pieces of stuff that exist today,
and I think we did that with the first stage. You’ve got the
Zenit heritage on the core, you’ve got these AJ-26 engines.
While they’re old, we know they work. At least they did back
then.
Eberly:
And they’re already built.
Pinkston:
And they’re already built. You don’t have to develop it,
you don’t have all this qualification to do. It’s simply
adapting, and that’s really how we’ve done business for
years.
Eberly:
I guess you could say we tend to be in a systems integrator role.
We’re the integrator. We buy the parts, we know what the parts
need to do. We do the interfaces, we do the analytical work, the simulations.
Zenit and Yuzhnoye really came through in terms of analyzing the parts.
We didn’t know anything about liquid oxygen in a tank, how much
boils off and how much pressure is created in the space that’s
not filled with liquid. That’s what they have expertise at,
how to keep that pressurized to keep the engines happy with enough
propellant so there’s no cavitation in the pumps.
They really came through for us, and that’s an area that we’ve
learned a lot from them. They brought their expertise that we really
didn’t have in-house to Orbital. We had never done this before.
The upper stack—after that clapping after the first stage separated,
that was our bread and butter for Orbital to fly the rest of the way.
I think we were able to effectively bring in these outside suppliers
with significant heritage.
Pinkston:
Still help us through the first stage.
Eberly:
They got us through the first stage, that’s right.
Hackler:
Just a question out of curiosity—you mentioned it was originally
called the Taurus II. Why was the name changed?
Eberly:
I think Taurus II was a study name that we had back in ’06,
and we just kept it. We had almost changed it a couple times, and
it originally was going to look more like a Taurus, which is kind
of a smaller solid rocket motor vehicle. Originally this was going
to have a smaller core, a three meter core and some strap-on solids,
so it was sort of like a Taurus. We did a trade study, and did away
with the solid rocket motor strap-on boosters and went with the 3.9-meter
core, which is bigger, and just have one configuration as our decision.
So it really started to look less and less like a Taurus.
One of the purposes for calling it Taurus II was we thought the first
launch of this rocket was going to come from Vandenberg Air Force
Base. They knew Taurus, and we thought that we’d have a better
time of it with the range safety folks if we called it Taurus II because
they would know instantly, “Oh, it’s Orbital Sciences
and it’s the same guys. We can approve their systems more readily.”
It turns out we haven’t been to Vandenberg, and we may not ever
go to Vandenberg to launch this thing. So all the reasons for keeping
it fell away, and our then CEO really thought this needed its own
name. Operationally we thought, “If I don’t change the
name now—.” Once you do a launch, you can never change
it. That’s really when he decided to do it.
Hackler:
The last question I have before I turn it back over to Rebecca Wright—one
of the goals of the COTS program is to encourage the space transportation
industry, not just for NASA’s purposes, but the broader commercial
sector. What other future uses of this new vehicle that you’ve
developed are you looking at?
Eberly:
We really want to launch spacecraft, provide a commercial launch service
to the government, as we do with our vehicles. Mainly to the government,
but also to commercial users. We want to launch spacecraft into low-Earth
orbit, or geosynchronous, or to Earth escape. We have on-ramped this
vehicle to the NASA Launch Services contract, which is a commercial
procurement vehicle. It’s for the government, but you do procurement,
you get an FAA launch license, and so on. In that sense it’s
commercial. We’ve also done the same with the Air Force contract,
the OSP-3, Orbital/Suborbital Program-3 contract.
Both of those are IDIQ, indefinite delivery, indefinite quantity-type
ordering schemes for the government. Basically, you get onto an approved
supplier list, and then when they have a mission that they want to
launch they issue a Request for Proposal, and you come in with your
best price. We’ve already gone through the process of writing
a long proposal, detailing all the services that you’ll provide,
and you have a statement of work. Now it’s just down to bidding.
You can’t win one of those contracts until your first launch.
Now that COTS has given us our first demonstration launch, we can
use that to bid on future spacecraft launches for the Air Force and
for NASA. That’s a direct benefit.
We also launched four secondaries on that test flight, and on these
next launches we hope to line up more secondary payloads to go fly
along and help us out a little bit. On the Orb-3 Mission, which will
be our fifth launch, there’s some extra capacity just for a
contractual reason. We’re flying an enhanced-capability rocket
with a standard-capability Cygnus, just because NASA ISS wanted certain
features on that Cygnus, dual-berth visiting vehicle features that
allow two visiting vehicles to be berthed at the same time. For unrelated
reasons we have some extra performance on that mission, so we’re
trying to use that up with a larger secondary payload.
Hackler:
Thank you.
Wright:
Was there anything else that y’all would like to add that we
haven’t covered? Anything about CRS?
Eberly:
We want to get past this development. We want to see Cygnus fly successfully
and berth with the Station, and that’ll be the last of the demonstrations.
Then it’s just on to supplying a regular cargo delivery service.
It’s going to be interesting to see that, and we just hope to
be a regular supplier, like space FedEx [Corporation, shipping service].
Just show up every now and then with some cargo for the astronauts
on schedule. That’s going to be our goal, to really make sure
that we can meet the dates going forward, now that we’re done
with development on the ground side and the rocket side.
That’s the challenge that Mike and I have, to productionize
this program and make sure our subs [subcontractors] are delivering
hardware that’s flightworthy on schedule, and we can get our
operations efficient enough that we can meet all these dates so we’re
not having to ask for any more delays. That’s painful for everyone,
and certainly the customer. That’s what the next phase is going
to be for CRS, providing that regular service to the Space Station.
We’re excited to be part of that and part of keeping the astronauts
working on the ISS.
Pinkston:
I think the whole idea of commercial resupply, it’s an exciting
direction that it’s going, shipping some of this away from big
government programs to commercial suppliers. Being on the front end
of that, you figure that’s a trend that’s going to go
for a long time. Just the way budgets are and will be going forward,
being on the front end of that trend is exciting.
We’re squarely focused on getting the COTS demonstration done.
I think back to since I’ve been on the program, all the significant
development milestones we’ve clicked off over the course of
the last seven, eight months, it’s pretty astounding. We’ve
got one more giant step to take with the COTS demo, and then it really
is on to making this a regular and repeatable service that we can
provide when it’s required. FedEx of the Space Station, and
we’re glad to do it.
Eberly:
I think to that point, NASA took some risk with this procurement.
They didn’t have the money to do a typical procurement where
they would write down every requirement and a long spec [specification]
on how you should design your rocket. Instead they gave us performance
requirements. “Deliver this cargo in the end, and we’ll
open the door for you at the ISS.” They still have to prescribe
the safety requirements, and that’s what Cygnus has to do, but
in terms of Antares there were no requirements that weren’t
internally derived by us or by the range that we had to go to.
NASA really could oversee us with an office of three people, and they
could bring in expertise as they needed. If you think about that and
the money they spent and the way they capped their risk, it was a
really successful way to develop a new aerospace system from the government’s
perspective. From our perspective, we’ll see. We’re still
trying to dig out of the hole that we’re in. We try to stay
out of sight of our CEO still, but he’ll get over it when we
get a few more launches going.
I think that’s the benefit. The aerospace industry is mature
enough to the point where you can say, “Go do this, deliver
something in orbit. I don’t have to tell you how to do it because
industry already knows how to do it.” I think that’s been
proven to be the case for SpaceX and for Orbital.
Wright:
We thank you both for your time and for all your great information.
We wish you great luck and we can’t wait until the end of summer
when we see the demonstration launch.
Pinkston:
Thank you.
Eberly:
Thank you.
[End
of Interview]
Return
to JSC Oral History Website